Deemed Undertaking. Fibrogen, Inc. v. Akebia Therapeutics, Inc.
In Fibrogen, Inc. v. Akebia Therapeutics, Inc. (Fed CA, 2022) the Federal Court of Appeal considered the federal version of the 'implied undertaking rule':
The implied undertaking rule. Amyotrophic Lateral Sclerosis Society of Essex County v. Windsor (City)
 I agree with FibroGen that the effect of Akebia’s motion is to circumvent the implied undertaking rule. The rule imposes an undertaking on the parties not to use information or documents that parties are compelled to produce in the course of a civil proceeding for any use other than that proceeding. The undertaking is a continuing obligation; it survives the end of the litigation and is only extinguished when the documents or answers are used in open court. The undertaking is made to the Court and is enforceable by motion. By bringing the motion after the end of the action under guise of seeking an order which would expand the scope of the documents that are "“public”" or "“non-public”", Akebia seeks to evade the obligations associated with the rule (Juman v. Doucette, 2008 SCC 8,  1 S.C.R. 157 at paras. 21, 27, 51 [Juman]; Duncan v. Lessing, 2018 BCCA 9, 5 B.C.L.R. (6th) 81 at para. 5 [Duncan]).
 The implied undertaking rule applies to both documentary and oral information obtained on discovery: such evidence is not to be used except for the purpose of that litigation unless and until the undertaking is varied by court order (Juman at para. 4) or until the documents are admitted into evidence and become part of the public court record. Whether documents produced or answers given are privileged and confidential is irrelevant to the undertaking (Juman at para. 27).
 A broader, purposeful understanding of what constitutes compulsion in the context of the implied undertaking rule has also been adopted by the Australian High Court. In Hearne v. Street,  HCA 36, (2008) 235 C.L.R. 125 at paragraph 96, the Court wrote:
Where one party to litigation is compelled, either by reason of a rule of court, or by reason of a specific order of the court, or otherwise, to disclose documents or information, the party obtaining the disclosure cannot, without the leave of the court, use it for any purpose other than that for which it was given unless it is received into evidence. The types of material disclosed to which this principle applies include documents inspected after discovery, answers to interrogatories, documents produced on subpoena, documents produced for the purposes of taxation of costs, documents produced pursuant to a direction from an arbitrator, documents seized pursuant to an Anton Piller order, witness statements served pursuant to a judicial direction and affidavits. [emphasis added] I also note that the appellate courts in British Columbia and Ontario, as well as the Federal Court, have framed the undertaking as applying to information which could not have otherwise been obtained by a legitimate means independent of the litigation (Duncan at para. 5; Andersen Consulting v. Canada, 2001 CanLII 22032 (FC),  2 F.C. 324, 199 F.T.R. 150 at para. 4, citing Goodman at 617; N.M. Paterson & Sons Ltd. v. St. Lawrence Seaway Management Corp., 2002 FCT 1247 (CanLII), 2002 F.C.T. 1247, 225 F.T.R. 308 at para. 4, aff’d 2004 FCA 210, 322 N.R. 83). As noted by Justice John B. Laskin in “The Implied Undertaking in Ontario” (1990), 11 Advocates Q. 298 at 312, there is no reason in principle why the rule should not apply to other forms of compulsion. This would apply to the witness statements of Dr. Guenzler–Pukall and Dr. Seeley.
 Akebia points to three cases in support of its position that the implied undertaking was not engaged as the preparation and disclosure of the statements was voluntary. I do not agree that these cases support that conclusion.
 Akebia relies on the decision of this Court in Canada v. Fio Corporation, 2015 FCA 236, 478 N.R. 194 [Fio]. The facts of Fio are markedly different. There, this Court found that the implied undertaking rule did not apply to voluntary disclosures made before the litigation commenced. The documents in question were voluntarily disclosed by the taxpayer to the Canada Revenue Agency during the course of an audit. There was no litigation pending.
 Likewise, the decision of the Manitoba Court of Appeal in J-Sons Inc. v. N. M. Paterson & Sons Limited, 2003 MBCA 156, 180 Man. R. (2d) 178 dealt with a voluntary, tactical decision to include an expert report in a pre-trial brief (at para. 21), which was then shared with counsel in Alberta conducting related litigation against the defendant. The implied undertaking was not breached, however, as the Manitoba rules of practice narrowed the broad scope of the common law rule to cover only certain, specified pre-trial disclosure. The rules did not require the pre-trial disclosure of the report, nor was there any order that did.
 Gilead Sciences, Inc. v. Teva Canada Limited, 2016 FC 31,  F.C.J. No. 378 [Gilead], discussed the circumstances where a party will be relieved from the undertaking, one of which is where the documents are sought to be used in a parallel action between the same parties on the same or similar issues. In the particular facts of Gilead, a party was relieved from the obligation where the undertaking was being relied on simply to delay access to documents in one proceeding which it knew would inevitably be part of the public record in a parallel proceeding. By insisting on adherence to the rule in these circumstances, the party was simply impeding the efficient administration of justice (at para. 19).
 I close on a point of practice and procedure. The Federal Court found that Akebia had no choice but to file the statements with the Court. This was not the case.
 In many cases, and this is one of them, a party seeking to be relieved from the implied undertaking does not need to file the documents in question. As noted by the British Columbia Court of Appeal in AM Gold Inc. v. Kaizen Discovery Inc., 2021 BCCA 70, 46 B.C.L.R. (6th) 135, it is often entirely unnecessary to include the information in the motion record in order to be relieved from the undertaking (at paras. 34-35). I agree with the observation of Groberman J.A. that "“[a] generic description of the circumstances, which does not disclose confidential information, will usually be sufficient to allow a court to determine whether the implied undertaking should be relaxed …”" (at para. 34).
In Amyotrophic Lateral Sclerosis Society of Essex County v. Windsor (City) (Ont CA, 2019) the Court of Appeal considered the deemed undertaking rule:
 The core purpose of the deemed undertaking rule is to protect the use to which the compelled production of a party’s private information can be put: Goodman v. Rossi (1995), 1995 CanLII 1888 (ON CA), 24 O.R. (3d) 359 (C.A.); Juman v. Doucette, 2008 SCC 8,  1 SCR 157. The opt-out information is not captured by that purpose. Rather, the opt-out information is information that relates to the identities of the class members who have agreed to be bound by the determinations made in the context of the class proceedings.
 Courts in Canada operate on the openness principle. A person who chooses to commence a court proceeding must do so publicly, subject only to exceptional circumstances where a pseudonym or initials may be used: C.G. v. Ontario (Health Insurance Plan General Manager), 2014 ONSC 5392, 327 O.A.C. 53 (Div. Ct.), at para. 7. The overarching principle applicable to all court proceedings is that the public is entitled to know what proceedings are commenced in our courts, along with the particulars of those proceedings, and whose rights are being submitted for adjudication. Any member of the public is entitled to have access to the court file, and to review all of the material that is filed with the court. In the same way, a defendant is entitled to know who is using the court process to advance a claim against it.
 The respondents, and the Divisional Court, rely on my endorsement in Markle v. Toronto (City),  O.J. No. 3024(S.C.J.) in support of their position. With respect, both mischaracterize my decision along with the nature of the information that was involved. In Markle, the information sought by the representative plaintiff was the private information of the City of Toronto. It was a list of the names and addresses of individuals who had been employed by, but were now retired from, the City. If the action proceeded to discovery, it would have been appropriate for counsel for the representative plaintiff to ask the City’s representative for a list of the retired persons whose claims would be covered by the class action. This request would be consistent with the requirement that rested on the City as a party, by virtue of s. 5(3) of the CPA, to provide its best information on the number of members of the proposed class as part of the certification process. Because Markle was being certified as a class action on consent, in order to ensure that notice of the class action was given in the most effective manner to the members of the class, I ordered that the City produce the detailed information in advance of the discovery process.
 In its reasons, the Divisional Court said, at para. 44, that “[t]he submission by the government defendants that the deemed undertaking applies only where a party is compelled to disclose its own private information in discovery” is inconsistent with my decision in Markle. To the contrary, that submission is entirely consistent with the decision, properly understood.
 As the class actions judge pointed out in his reasons, at para. 8:
As a result of an application under the Municipal Freedom of Information and Protection of Privacy Act, R.S.O. 1990, c. M56 (“MFIPP”) a list of all 850+ potential class members is publicly available on more than one web site. Underlying this appeal is an evident concern that the appellants might misuse the opt-out information because of the earlier issues surrounding the opt-out process. The class action judge declined to continue the earlier protection order he had made arising out of those earlier issues, although in so doing he applied the stricter test for the granting of a sealing order dictated by Sierra, rather than determining whether, at that stage, a confidentiality order should be made pursuant to s. 12 of the CPA to preserve the integrity of the class proceedings.
All that the opt-out information does is permit someone to reduce the publicly known 850+ potential class members to the actual class members. Both the public and the appellants are entitled to know the identities of the class members who are advancing the claims unless a confidentiality order is made pursuant to s. 12 of the CPA.
 In any event, to the degree that any such concern might arise, s. 12 of the CPA gives a class action judge all of the authority that he or she needs to address that issue. Indeed, the section could not be more broadly worded. Section 12 reads:
The court, on the motion of a party or class member, may make any order it considers appropriate respecting the conduct of a class proceeding to ensure its fair and expeditious determination and, for the purpose, may impose such terms on the parties as it considers appropriate. If such a concern does arise, s. 12 is the proper route to be used to address it. There is no doubt that this class actions judge was well aware of the authority he had under that section as he expressly referred to it in his reasons dealing with the issues that arose over the original opt-out period. To the degree that the respondents feel that the class actions judge did not turn his mind to using that authority to address their expressed concerns regarding the lifting of the protection order, there is nothing that prevents the respondents from returning before the class actions judge to seek a remedy. That is the proper approach to take. The plain wording of r. 31.01.1 should not be distorted in order to effect such a remedy.